This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Shell wrappers and GCC - FYA
- To: Bruce Korb <bkorb at veritas dot com>
- Subject: Re: Shell wrappers and GCC - FYA
- From: "Zack Weinberg" <zackw at Stanford dot EDU>
- Date: Thu, 28 Jun 2001 18:28:14 -0700
- Cc: Neil Booth <neil at daikokuya dot demon dot co dot uk>, Bruce Korb <bkorb at pacbell dot net>, gcc at gcc dot gnu dot org
On Wed, Jun 27, 2001 at 01:19:07PM -0700, Bruce Korb wrote:
>
> As everyone knows, GCC is _already_ often wrapped inside
> a 200KB shell script -- libtool. As it happens, I am
> trying to convert that to a binary wrapper a la fixincludes,
> but, for the time being, .... just for your amusement.
I believe everyone knows what I think of libtool, already.
I'm not sure whether libtool is an existence proof that you _can_
write a shell script that handles its arguments correctly, or a
demonstration that you may try but you are doomed to failure. (How
well does it cope with arguments with shell special characters in
them?)
In any case, arguments about the performance hit of shell wrappers are
only underscored by libtool. And if g++ (say) were to be a shell
wrapper, when used with libtool, that would be TWO shell wrappers.
One of my deep-future projects is to get Make to run cc1 directly,
cutting out all the overhead associated with the driver. There'd
still be a chunk of code executing in Make's process that did the
argument munging, of course. But with Neil's new driver that should
be getting down to the point where it is comparably expensive to the
startup overhead...
--
zw The master came to a yatai which was selling hot dogs. 'What do you
want on your hot dog?' he was asked. 'Nothing,' he replied. Then the
hot dog was enlightened.
-- a koan by Eric Hallstrom